Sign up for our newsletters to get the latest updates and exclusive content on industry-leading AI news.
Substrate, a startup founded by tech industry experts Rob Cheung and Ben Guo, recently came out of stealth mode, introducing an advanced AI development platform. The company also secured $8 million in funding from Lightspeed Venture Partners to expand its team and enhance its product offerings.
Substrate’s mission is to make AI more accessible by offering a unified platform for businesses to build, deploy, and manage machine learning models and workflows. Their main product is an API that lets developers create sophisticated AI processes by combining high-quality, open-source models, all optimized and curated by Substrate.
The company believes its platform will make it easier and more affordable for businesses to leverage advanced AI like large language models (LLMs) and other generative AI techniques. This could speed up AI adoption in various sectors, including content creation, business analytics, and customer support.
Breaking down complex problems into manageable pieces is key. Rob Cheung, Substrate’s co-founder and CEO, explained that the current challenges with integrating AI, especially LLMs, are accuracy, cost, and latency. Substrate tackles these issues by allowing developers to fragment large problems into smaller, more manageable ones.
Cheung likened it to Google Search, which uses 15-20 different machine learning models together. If you rely on a single massive prompt, you might get good results only 10% of the time, and be uncertain the rest of the time. By breaking down the problem and managing it effectively, accuracy improves significantly.
While big companies like Google have the infrastructure to manage thousands of ML models, most businesses don’t. Centralizing performance optimization and offering it as a service is Substrate’s solution. Cheung mentioned that one of their big clients, Substack, isn’t interested in handling ML infrastructure. They just want easy-to-use components to build their ML workloads.
Ben Guo, Substrate’s co-founder, noted that early customers like Substack, who used the platform to create summaries and categorize blog posts, have seen great benefits. This platform allows them to run all their models in one place, improving speed, cutting costs, and boosting reliability.
Guo also highlighted that many enterprises don’t want to sift through the ever-growing list of open-source AI models. Substrate’s curated models help by selecting the most effective ones available. Additionally, Substrate is focused on improving the developer experience by providing simple templates for common enterprise use cases, which Guo likens to his experience at Stripe, where simplifying integrations provided significant value.
With powerful and accessible AI tools like large language models, Substrate’s platform can help businesses turn these capabilities into real-world applications. Substrate aims to simplify and economize AI development, much like how cloud platforms revolutionized general-purpose computing.
The $8 million funding will help Substrate grow its platform, team, and market presence to reach more businesses. With experienced founders and early success, Substrate is poised to become a significant player in enterprise AI.
Stay updated with the latest news by subscribing to our daily newsletter.